Note: ‘*’ represents equal contribution or alphabetical order.
Preprints
[P1] Learning to Reason with LLMs. OpenAI 2024. (core contributor to the RL algorithms).
[P2] GPT-4o mini: advancing cost-efficient intelligence. OpenAI 2024. (core contributor to the optimization of the model).
[P3] GPT-4o. OpenAI 2024. (minor contributions to optimization and evals).
[P4] On the Opportunities and Risks of Foundation Models. Rishi Bommasani, Drew A. Hudson, …, Ananya Kumar, …, Percy Liang (116 authors, alphabetical within ellipses).
Conference Papers
[C21] Generative Classifiers Avoid Shortcut Solutions. Alexander C. Li, Ananya Kumar, Deepak Pathak. International Conference on Learning Representations (ICLR) 2025.
[C20] How to Fine-Tune Vision Models with SGD. Ananya Kumar, Ruoqi Shen, Sebastien Bubeck, Suriya Gunasekar. International Conference on Learning Representations (ICLR) 2024.
[C19] Conservative Prediction via Data-Driven Confidence Minimization. Caroline Choi*, Fahim Tajwar*, Yoonho Lee, Huaxiu Yao, Ananya Kumar, Chelsea Finn. Transactions on Machine Learning Research (TMLR) 2024.
[C18] Holistic Evaluation of Language Models. Percy Liang*, Rishi Bommasani*, Tony Lee*, Dimitris Tsipras, Dilara Soylu, Michihiro Yasunaga, Yian Zhang, Deepak Narayanan, Yuhuai Wu, Ananya Kumar, … (Ananya led the section on Uncertainty Calibration). Transactions on Machine Learning Research (TLMR Featured) 2023.
[C17] Are Sample-Efficient NLP Models More Robust? Nelson Liu, Ananya Kumar, Percy Liang, Robin Jia. Annual Meeting of the Association for Computational Linguistics (ACL) 2023.
[C16] Finetune like you pretrain: Improved finetuning of zero-shot vision models. Sachin Goyal, Ananya Kumar, Sankalp Garg, Zico Kolter, Aditi Raghunathan. Conference on Computer Vision and Pattern Recognition (CVPR) 2023.
[C15] Surgical Fine-Tuning Improves Adaptation to Distribution Shifts. Yoonho Lee*, Annie S. Chen*, Fahim Tajwar, Ananya Kumar, Huaxiu Yao, Percy Liang, Chelsea Finn. International Conference on Learning Representations (ICLR) 2023.
[C14] Picking on the Same Person: Does Algorithmic Monoculture lead to Outcome Homogenization? Rishi Bommasani, Kathleen A. Creel, Ananya Kumar, Dan Jurafsky, Percy Liang. Neural Information Processing Systems (NeurIPS) 2022 and ACM Conference on Equity and Access in Algorithms, Mechanisms, and Optimization (EEAMO) 2022.
[C13] Beyond Separability: Analyzing the Linear Transferability of Contrastive Representations to Related Subpopulations. Jeff Z. HaoChen, Colin Wei, Ananya Kumar, Tengyu Ma. Neural Information Processing Systems (NeurIPS) 2022.
[C12] Connect, Not Collapse: Explaining Contrastive Learning for Unsupervised Domain Adaptation. Kendrick Shen*, Robbie Jones*, Ananya Kumar*, Sang Michael Xie*, Jeff Z. HaoChen, Tengyu Ma, Percy Liang. International Conference on Machine Learning (ICML Long Talk) 2022. 2.1% long talk acceptance rate.
[C11] Calibrated ensembles can mitigate accuracy tradeoffs under distribution shift. Ananya Kumar, Tengyu Ma, Percy Liang, Aditi Raghunathan. Conference on Uncertainty in Artificial Intelligence (UAI) 2022.
[C10] Fine-Tuning can Distort Pretrained Features and Underperform Out-of-Distribution. Ananya Kumar, Aditi Raghunathan, Robbie Jones, Tengyu Ma, Percy Liang. International Conference on Learning Representations (ICLR Oral) 2022. 1.6% oral acceptance rate.
[C9] Extending the WILDS benchmark for unsupervised adaptation. Shiori Sagawa*, Pang Wei Koh*, Tony Lee*, Irena Gao*, Sang Michael Xie, Kendrick Shen, Ananya Kumar, Weihua Hu, Michihiro Yasunaga, Henrik Marklund, Sara Beery, Etienne David, Ian Stavness, Wei Guo, Jure Leskovec, Kate Saenko, Tatsunori Hashimoto, Sergey Levine, Chelsea Finn, and Percy Liang. International Conference on Learning Representations (ICLR Oral) 2022. 1.6% oral acceptance rate.
[C8] In-N-Out: Pre-Training and Self-Training using Auxiliary Information for Out-of-Distribution Robustness. Sang Michael Xie*, Ananya Kumar*, Robbie Jones*, Fereshte Khani, Tengyu Ma, Percy Liang. International Conference on Learning Representations (ICLR) 2021.
[C7] Selective classification can magnify disparities across groups. Erik Jones*, Shiori Sagawa*, Pang Wei Koh*, Ananya Kumar, and Percy Liang. International Conference on Learning Representations (ICLR) 2021.
[C6] Self-Training Avoids Using Spurious Features Under Domain Shift. Yining Chen*, Colin Wei*, Ananya Kumar, Tengyu Ma. Neural Information Processing Systems (NeurIPS) 2020.
[C5] Understanding Self-Training for Gradual Domain Adaptation. Ananya Kumar, Tengyu Ma, Percy Liang. International Conference on Machine Learning (ICML) 2020.
[C4] Verified Uncertainty Calibration. Ananya Kumar, Percy Liang, Tengyu Ma. Neural Information Processing Systems (NeurIPS Spotlight) 2019. 3.0% spotlight / oral acceptance rate.
[C3] Rigorous Agent Evaluation: An Adversarial Approach to Uncover Catastrophic Failures. Jonathan Uesato*, Ananya Kumar*, Csaba Szepesvari*, Tom Erez, Avraham Ruderman, Keith Anderson, Krishnamurthy (Dj) Dvijotham, Nicolas Heess, Pushmeet Kohli. International Conference on Learning Representations (ICLR) 2019.
[C2] Approximate Convex Hull of Data Streams. Avrim Blum*, Vladimir Braverman*, Ananya Kumar*, Harry Lang*, Lin Yang* (Alphabetical Order). International Colloquium on Automata, Languages, and Programming (ICALP) 2018.
[C1] Parallel Functional Arrays. Ananya Kumar, Guy E. Blelloch and Robert Harper. Principles of Programming Languages (POPL) 2017.
Workshop Papers
[–] Calibrated Ensembles – a Simple Way to Mitigate ID-OOD Accuracy Tradeoffs. Ananya Kumar, Aditi Raghunathan, Tengyu Ma, Percy Liang. NeurIPS DistShift Workshop.
[W3] No True State-of-the-Art? OOD Detection Methods are Inconsistent across Datasets. Fahim Tajwar, Ananya Kumar*, Sang Michael Xie*, Percy Liang. ICML UDL Workshop.
[W2] Uncovering Surprising Behaviors in Reinforcement Learning via Worst-Case Analysis. Avraham Ruderman, Richard Everett, Bristy Sikder, Hubert Soyer, Charles Beattie, Jonathan Uesato, Ananya Kumar and Pushmeet Kohli. Oral presentation in ICLR SafeML Workshop 2019. 4.8% oral acceptance rate.
[–] Rigorous Agent Evaluation: An Adversarial Approach to Uncover Catastrophic Failures. Jonathan Uesato*, Ananya Kumar*, Csaba Szepesvari*, Tom Erez, Avraham Ruderman, Keith Anderson, Krishnamurthy (Dj) Dvijotham, Nicolas Heess, Pushmeet Kohli. Oral presentation in NeurIPS Workshop on Security in Machine Learning 2018. 4.6% oral acceptance rate.
[W1] Consistent Generative Query Networks. Ananya Kumar, S. M. Ali Eslami, Danilo Rezende, Marta Garnelo, Fabio Viola, Edward Lockhart, Murray Shanahan. NeurIPS Workshop on Bayesian Deep Learning 2018 and NeurIPS Physics Workshop 2018.
[–] Approximate Convex Hull of Data Streams. Avrim Blum*, Vladimir Braverman*, Ananya Kumar*, Harry Lang*, Lin Yang* (Alphabetical Order). Fall Workshop on Computational Geometry (FWCG) 2017.
Ananya Kumar 